Date of Award

5-2018

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Civil Engineering

Committee Member

Dr. Bradley J Putman, Committee Chair

Committee Member

Dr. Prasad Rao Rangaraju

Committee Member

Dr. Amir Poursaee

Abstract

Characterization of internal structure is necessary to understand the behavior of construction materials under different conditions. Image analysis possesses the technological advantages to better understand construction materials and it also provides information that can help to improve the properties of various construction materials. The purpose of this study was to evaluate the internal structure of various construction materials using image analysis and investigate the infiltration in porous friction course mixtures using thermal image analysis. Automated image processing algorithms were developed to determine the a) long-term binder draindown in asphalt mixtures over time, b) roughness of asphalt mixtures, c) fiber distribution in ultra-high performance concrete mixtures, and d) amount of unreacted ground glass fiber (GGF) particles in geopolymers. Although, the image analysis results did not compare numerically with measured results in the first two studies (a and b), the accuracy can be improved through calibration. In the other two studies (c and d), image analysis provided accurate measurements of the desired information, which can be used to help alter the composition of the materials to improve the performance. Results also helped to analyze the internal structure thus improving the ability to understand the behavior of construction materials. To simulate and monitor difficult to detect observations in full scale, a study was also conducted on pervious concrete friction course mixtures to evaluate infiltration and lateral internal flow of water using thermal image analysis. The internal spread of infiltrated water was detected using thermal image analysis and the effect of slope and thickness on infiltration was analyzed based on various parameters. Results were shown to be repeatable and enhanced the importance of using this non-destructive, cost effective and a time efficient approach.

Share

COinS